WMT08

Workshop on Machine Translation


The Third Workshop on Machine Translation (WMT08) took place on 19 June, 2008, in Columbus, Ohio. It was organised by WMT.

statmt.org/wmt08/program.html

Table of contents
  1. Tasks
  2. Schedule
  3. Results
    1. Translation
      1. → English
      2. English →

Tasks

Schedule

   
08:40 – 08:50 Opening Remarks
08:50 – 09:10 Session 1: Full Papers
An Empirical Study in Source Word Deletion for Phrase-Based Statistical Machine Translation
Chi-Ho Li, Hailei Zhang, Dongdong Zhang, Mu Li, Ming Zhou
09:10 – 09:30 Rich Source-Side Context for Statistical Machine Translation
Kevin Gimpel, Noah A. Smith
09:30 – 09:50 Discriminative Word Alignment via Alignment Matrix Modeling
Jan Niehues, Stephan Vogel
09:50 – 10:10 Regularization and Search for Minimum Error Rate Training
Daniel Cer, Daniel Jurafsky, Christopher Manning
10:10 – 10:30 Learning Performance of a Machine Translation System: a Statistical and Computational Analysis
Marco Turchi, Tijl De Bie, Nello Cristianini
10:30 – 11:00 ☕️
11:00 – 11:20 Session 2: Full Papers
Using Syntax to Improve Word Alignment Precision for Syntax-Based Machine Translation
Victoria Fossum, Kevin Knight, Steven Abney
11:20 – 11:40 Using Shallow Syntax Information to Improve Word Alignment and Reordering for SMT
Josep M. Crego, Nizar Habash
11:40 – 12:00 Improved Tree-to-String Transducer for Machine Translation
Ding Liu, Daniel Gildea
12:00 - 12:40 Invited Talk
Daniel Marcu
12:40 - 14:00 🍴
14:00 - 14:30 Session 3: Shared Task
Further Meta-Evaluation of Machine Translation
Chris Callison-Burch, Cameron Fordyce, Philipp Koehn, Christof Monz, Josh Schroeder
14:30 – 14:40 Limsi’s Statistical Translation Systems for WMT08
Daniel Déchelotte, Gilles Adda, Alexandre Allauzen, Hélène Bonneau-Maynard, Olivier Galibert, Jean-Luc Gauvain, Philippe Langlais, François Yvon
14:40 – 14:50 The MetaMorpho Translation System
Attila Novák, László Tihanyi, Gábor Prószéky
14:50 – 15:00 Meteor, M-BLEU and M-TER: Evaluation Metrics for High-Correlation with Human Rankings of Machine Translation Output
Abhaya Agarwal, Alon Lavie
15:00 - 15:30 Booster Session: Shared Task
  Shared Translaton Task
  First Steps towards a General Purpose French/English Statistical Machine Translation System
Holger Schwenk, Jean-Baptiste Fouet, Jean Senellart
  The University of Washington Machine Translation System for ACL WMT 2008
Amittai Axelrod, Mei Yang, Kevin Duh, Katrin Kirchhoff
  The TALP-UPC Ngram-Based Statistical Machine Translation System for ACL-WMT 2008
Maxim Khalilov, Adolfo Hernández H., Marta R. Costa-jussà, Josep M. Crego, Carlos A. Henríquez Q., Patrik Lambert, José A. R. Fonollosa, José B. Mariño, Rafael E. Banchs
  European Language Translation with Weighted Finite State Transducers: The CUED MT System for the 2008 ACL Workshop on SMT
Graeme Blackwood, Adrià de Gispert, Jamie Brunning, William Byrne
  Effects of Morphological Analysis in Translation between German and English
Sara Stymne, Maria Holmqvist, Lars Ahrenberg
  Towards better Machine Translation Quality for the German-English Language Pairs
Philipp Koehn, Abhishek Arun, Hieu Hoang
  Phrase-Based and Deep Syntactic English-to-Czech Statistical Machine Translation
Ondrej Bojar, Jan Hajic
  Improving English-Spanish Statistical Machine Translation: Experiments in Domain Adaptation, Sentence Paraphrasing, Tokenization, and Recasing
Preslav Nakov
  Improving Word Alignment with Language Model Based Confidence Scores
Nguyen Bach, Qin Gao, Stephan Vogel
  Kernel Regression Framework for Machine Translation: UCL System Description for WMT 2008 Shared Translation Task
Zhuoran Wang, John Shawe-Taylor
  Using Syntactic Coupling Features for Discriminating Phrase-Based Translations (WMT-08 Shared Translation Task)
Vassilina Nikoulina, Marc Dymetman
  Statistical Transfer Systems for French-English and German-English Machine Translation
Greg Hanneman, Edmund Huber, Abhaya Agarwal, Vamshi Ambati, Alok Parlikar, Erik Peterson, Alon Lavie
  TectoMT: Highly Modular MT System with Tectogrammatics Used as Transfer Layer
Zdenek Zabokrtsky, Jan Ptacek, Petr Pajas
  MaTrEx: The DCU MT System for WMT 2008
John Tinsley, Yanjun Ma, Sylwia Ozdowska, Andy Way
  Can we Relearn an RBMT System?
Loïc Dugast, Jean Senellart, Philipp Koehn
  Using Moses to Integrate Multiple Rule-Based Machine Translation Engines into a Hybrid System
Andreas Eisele, Christian Federmann, Hervé Saint-Amand, Michael Jellinghaus, Teresa Herrmann, Yu Chen
  Incremental Hypothesis Alignment for Building Confusion Networks with Application to Machine Translation System Combination
Antti-Veikko Rosti, Bing Zhang, Spyros Matsoukas, Richard Schwartz
  Shared Evaluation Task
  The Role of Pseudo References in MT Evaluation
Joshua Albrecht, Rebecca Hwa
  Ranking vs. Regression in Machine Translation Evaluation
Kevin Duh
  A Smorgasbord of Features for Automatic MT Evaluation
Jesus Gimenez, Lluis Marquez
15:30 - 16:40 Poster Session
☕️
16:40 – 17:00 Session 4: Full Papers
Fast, Easy, and Cheap: Construction of Statistical Machine Translation Models with MapReduce
Chris Dyer, Aaron Cordova, Alex Mont, Jimmy Lin
17:00 – 17:20 Dynamic Model Interpolation for Statistical Machine Translation
Andrew Finch, Eiichiro Sumita
17:20 – 17:40 Improved Statistical Machine Translation by Multiple Chinese Word Segmentation
Ruiqiang Zhang, Keiji Yasuda, Eiichiro Sumita
17:40 – 18:00 Optimizing Chinese Word Segmentation for Machine Translation Performance
Pi-Chuan Chang, Michel Galley, Christopher Manning

Results

Full results of the shared tasks: Further Meta-Evaluation of Machine Translation

Translation

The results were determined with a relative ranking, the ≥ others (“greater than or equal to others”) score, a constituent ranking, and a yes or no constituent judgement.

→ English

Language pair System ≥ others Constituent Acceptability
Czech Commentary → DCU 0.745 - -
Czech News → DCU 0.675 - -
French Europarl → LIUM-SYSTRAN
UEDIN
0.792
-
0.917
-
-
0.910
French News → LIUM-SYS-CONTR
LIMSI
0.693
-
-
0.801
-
0.805
German Europarl → UEDIN
SAAR
0.679
-
-
0.843
0.856
-
German News → RBMT2
SAAR-CONTR
RBMT4
0.742
-
-
-
0.775
-
-
-
0.738
Spanish Europarl → LIMSI
SAAR
0.780
-
-
0.893
0.902
-
Spanish News → RBMT4
CUED-CONTR
UPC
0.768
-
-
-
0.818
-
-
-
0.707
Hungarian News → MORPHOLOGIC 0.853 - -

English →

Language pair System ≥ others Constituent Acceptability
→ Czech Commentary CU-BOJAR 0.714 0.732 0.594
→ Czech News PC-TRANSLATOR
CU-BOJAR
0.715
-
0.674
-
-
0.540
→ French Europarl LIMSI
LIUM-SYSTRAN
0.791
-
-
0.881
-
0.843
→ French News RBMT4
LIUM-SYSTRAN
0.780
-
0.768
-
-
0.748
→ German Europarl RBMT4
UEDIN
0.681
-
-
0.822
-
0.849
→ German News RBMT2
RBMT4
0.787
-
-
0.748
-
0.750
→ Spanish Europarl LIMSI
UPC
0.737
-
0.855
-
-
0.903
→ Spanish News RBMT5
RBMT4
0.724
-
-
0.754
-
0.759

Want to learn more about WMT08?


Edit this article →

Machine Translate is created and edited by contributors like you!

Learn more about contributing →

Licensed under CC-BY-SA-4.0.

Cite this article →